A model for integrating cloud and fog computing for IoT applications

Abstract

Although Fog computing solves some of the challenges faced by implementing some IoT applications or cyber-physical systems (CPS), Fog nodes may not be able to meet the performance, energy, and latency constraints of future IoT or CPS applications unless the Fog computing architecture is tailored to meet these application requirements. This adaptation is required at both the system and node levels for Fog computing.
This paper describes an integrated cloud and fog architecture with IoT that enables increased performance, energy efficiency, reduced latency, and scalability for future IoT applications. Fog nodes (e.g., edge servers, smart routers, base stations) receive computing process requests for sensed data from various IoT devices. To improve application performance, energy efficiency, and real-time responsiveness, the design proposes a reconfigurable Fog node architecture that analyzes application characteristics and reconfigures architectural resources to better meet peak load demands. In particular, the focus is on a reconfigurable hardware layer, in which FPGA devices are used to provide the hardware resilience features required by the applications of these systems.